 |
|||||
|
|
|
|
|
||
|
|
|||||
|
|
|
|
|
||
The defendant and the government assert that the rule of Miranda v. Arizona, 384 U. S. 436 (1966) must be required by the Constitution, because otherwise it could not be binding on state courts. See Brief for the United States 23-24; Brief for Petitioner 20-21. They assert this with all the confidence of a bridge player who has just led the ace of trumps for the rubber trick, and thus gloss over the main point of this case.
Their ace is actually a deuce. There are many areas of law in which this Court establishes federal common law (i.e., judge-made) rules binding on the states yet subject to revision or abrogation by Congress. Miranda is just one of many.
The theory of the legitimacy of such rules is laid out in Monaghan, Foreword: Constitutional Common Law, 89 Harv. L. Rev. 1 (1975). The defense lawyer amici refer to this work disparagingly as a "twenty-five-year-old law review article," see Brief for National Ass'n of Criminal Defense Lawyers et al. as Amici Curiae 16, as if the force of the argument is somehow diminished merely by the passage of time. Wisdom too often never comes, and so one ought not reject it merely because it came early. Cf. Henslee v. Union Planters Bank, 335 U. S. 595, 600 (1949) (Frankfurter, J., dissenting).
"The clarion yet careful pronouncement of Erie, 'There is no federal general common law,' opened the way to what, for want of a better term, we may call specialized federal common law." H. Friendly, Benchmarks 178 (1967). There are many areas which this Court has held to be governed by federal law even though that law cannot be found in the Constitution or any Act of Congress. Admiralty and maritime law is one such area. The Constitution includes these cases in the jurisdiction of federal courts. See U. S. Const., Art. III, § 2. However, the courts of one sovereign may hear cases arising under the laws of another, see The Federalist No. 82, p. 493 (C. Rossiter ed. 1961) (A. Hamilton), so it does not follow that the substance of maritime law must be federal. Even so, this Court does hold that the law is federal, see, e.g., Moragne v. State Maritime Lines, Inc., 398 U. S. 375, 401, n. 15 (1970), fashioned by courts in the absence of congressional action but always subject to revision by Congress. See In re Garnett, 141 U. S. 1, 12 (1891); 1 R. Rotunda & J. Nowak, Treatise on Constitutional Law § 3.4, pp. 352-355 (3d ed. 1999).
Federal common law, and not the state law of contracts and shipping documents, determined the outcome in Banco Nacional de Cuba v. Sabbatino, 376 U. S. 398 (1964). In a diversity jurisdiction case, see id., at 421, n. 20, the federal common law "act of state doctrine" was held controlling. See id., at 439. Although it has "constitutional underpinnings," this doctrine is not constitutionally required, see id., at 423, yet it determined the outcome of a case otherwise governed by state law.
Labor contracts are another area where federal common law has trumped state law. Congress has assigned labor contract cases to federal courts, 29 U. S. C. §185(a), and this statute, purely jurisdictional on its face, has been held to "express[ ] a federal policy that the substantive law to apply . . . 'is federal law, which the courts must fashion from the policy of our national labor laws.' " Allis-Chalmers Corp. v. Lueck, 471 U. S. 202, 209 (1985) (quoting Textile Makers v. Lincoln Mills, 353 U. S. 448, 456 (1957)). Under this doctrine, Allis-Chalmers reversed the judgment of a state court in a state-law tort case, on the ground that allowing the tort claim would subvert the congressional goal of a unified body of labor contract law. Id., at 220.
Federal interests similarly scuttled a state-law tort suit in Boyle v. United Technologies Corp., 487 U. S. 500 (1988). No federal statute proscribed this design defect suit against a federal contractor.
"But we have held that a few areas, involving 'uniquely federal interests,' [citation] are so committed by the Constitution and laws of the United States to federal control that state law is pre-empted and replaced, where necessary, by federal law of a content prescribed (absent explicit statutory directive) by the courts--so-called 'federal common law.' " Id., at 504 (citing Banco Nacional, supra, and other cases).
The existence of federal common law binding on state courts is not an "untenable theory," cf. Brief for the United States 26, but rather a legal doctrine established beyond serious dispute. The government's catalog of prophylactic rules, see id., at 44-47, confirms rather than refutes that criminal procedure cases are included.
Those seeking to discredit the federal common law thesis as applied to criminal procedure have claimed it is limited to cases where uniformity is the federal interest. See Schrock & Welsh, Reconsidering the Constitutional Common Law, 91 Harv. L. Rev. 1117, 1134 (1978). Boyle negates this notion. The federal interest threatened by the state tort law was the ability of the government to get its contracts filled at the lowest possible price, Boyle, supra, 487 U. S., at 507, not uniformity of law. The touchstone is "uniquely federal interests," id., at 504, which might be a need for uniformity, see id., at 508, or might be something else.
That "something else" may be enforcement of federal rights, whether constitutional or statutory. In Town of Newton v. Rumery, 480 U. S. 386, 390-391 (1987), the plaintiff had entered into a contract releasing his federal civil rights claims, primarily Fourth Amendment claims, against the town. The Court held perfunctorily that the validity of the release would be governed by federal law. Id., at 392. That law was not established by statute. It was determined by courts "by reference to traditional common-law principles." Ibid.
Considering Town of Newton in light of Boyle, we may ask what is the "uniquely federal interest" that allows federal common law to override otherwise applicable state law on the validity of contracts. Only one interest appears pertinent--the federal interest in enforcing the Fourteenth Amendment, including the "incorporated" provisions of the Bill of Rights. That is precisely the same interest involved in the remedies and criminal procedure cases.
When one person has wrongfully injured another in his person, property, or reputation, the usual remedy is damages in tort in a suit brought under state law. However, when the malefactor is an employee of the federal government and the wrong is a violation of the Constitution, the plaintiff will find his remedy governed largely by federal law, primarily judge-made federal law.
On one side, the plaintiff who seeks traditional tort remedies will find the reach of the state law remedy limited by the doctrine of immunity. In Barr v. Matteo, 360 U. S. 564 (1959) and Howard v. Lyons, 360 U. S. 593 (1959), the plaintiffs brought libel suits under local defamation law. See Barr, at 577, n. * (Black, J., concurring in the judgment). Both suits were barred by a body of federal law, which "has in large part been of judicial making." Id., at 569 (plurality opinion). This federal common law prevailed over state and District of Columbia law, but it was subject to the power of Congress to substitute a different rule. See id., at 577 (Black, J., concurring in the judgment).
On the other side of the remedial coin, we find Bivens v. Six Unknown Fed. Narcotics Agents, 403 U. S. 388 (1971). That case created, in effect, a federal common law tort action for the violation of federal constitutional rights by federal officers. The Bivens action, however, is not constitutionally required. Congress has broad powers over the scope of remedies. In Schweiker v. Chilicky, 487 U. S. 412 (1988), the Court declined to extend Bivens to a case of denial of a benefit without due process because Congress had provided another remedy, even though that remedy fell well short of complete relief. Id., at 425. "It must be plain that Congress necessarily has a wide choice in the selection of remedies, and that a complaint about action of this kind can rarely be of constitutional dimension." Hart, The Power of Congress to Limit the Jurisdiction of Federal Courts: An Exercise in Dialectic, 66 Harv. L. Rev. 1362, 1366 (1953).
Given these rules, we may ask several questions. By what authority does the United States Supreme Court block a state-law tort action when no federal statute or constitutional provision forbids it? By what authority does the Court authorize a federal-question lawsuit by one individual against another in the absence of a statute providing an action? And by what authority does Congress substitute an incomplete remedy for that cause of action?
The answer must be that protection of federal constitutional rights and protection of federal employees from lawsuits over the performance of their duties are "uniquely federal interests" within the meaning of Boyle, supra. Courts create the governing law, but Congress has the last word if it chooses to speak. Either way, federal law prevails over state law to the extent of a conflict. See R. Fallon, D. Meltzer, & D. Shapiro, Hart & Wechsler's The Federal Courts and the Federal System 876-877 (4th ed. 1996).
The rule of Chapman v. California, 386 U. S. 18 (1967) is analogous to Bivens. See Meltzer, Harmless Error and Constitutional Remedies, 61 U. Chi. L. Rev. 1, 29-30 (1994). A criminal appellant with a constitutional claim, like a Bivens plaintiff, is the moving party, claims that his rights have been violated, claims he has been injured by the violation, and seeks relief. The injury and relief components of the claim are where federal common law comes in.
Chapman involved a state constitutional provision, the predecessor of Cal. Const., Art. VI, §13, which established a standard for deciding reversible versus harmless error. See Chapman, supra, 386 U. S., at 20, n. 3. Chapman held that federal law overrode the state standard. Id., at 20-21. "The Chapman opinion was cryptic, however, about the source of the rule it announced." Meltzer, supra, 61 U. Chi. L. Rev., at 2. Space does not permit an extended discussion here, but Professor Meltzer reviews the possible sources and concludes "that the harmless error rule should be seen as constitutional common law." Id., at 26. In particular, the Chapman standard is not mandated by the Constitution. See id., at 24-26. As the Chapman opinion itself states, the responsibility, and presumably the authority, of the Court to fashion a rule exists only "in the absence of appropriate congressional action." 386 U. S., at 21. A court-made rule which overrides a state constitutional provision, yet is subject to revision by Congress, can only be explained as federal common law.
Miranda and Chapman are strongly analogous in that both rules recognize the inherent uncertainty of the underlying determination, and both are based on an assessment of the relative harm of erring in one direction versus the other. See supra, at 6. If we always knew to a certainty whether an error contributed to a verdict or not, there would be no need for a harmless error "standard." Errors known to be harmful would be remedied by reversal, and those known to be harmless would not. A standard of "beyond a reasonable doubt" implements a value judgment that error in one direction is far worse than error in the other. See In re Winship, 397 U. S. 358, 364 (1970). The Chapman standard accepts reversal of a large number of judgments for errors which probably did not affect the verdict as the price for fixing those errors that did affect the verdict, but which might be erroneously judged harmless under a lesser standard. That reasonable people might differ on this value judgment is illustrated by the large number of states with less rigid rules before Chapman. See Meltzer, supra, 61 U. Chi. L. Rev., at 22, n. 89. Miranda similarly skews the risk of error. A large number of voluntary, and hence constitutional, confessions are excluded in order to minimize the risk of erroneously letting in an occasional involuntary one. See supra, at 6-7. Miranda and Chapman belong to the same species. Both are federal common law rules, created to safeguard constitutional rights but not themselves constitutionally required.
The objection is made that recognizing such a common-law-making power in this Court would dangerously expand the Court's power. See Schrock and Welsh, supra, 91 Harv. L. Rev., at 1126-1127. There are two answers. First, the power should be carefully limited to that needed to effectuate constitutional rights and not to expand the scope of those rights, similar to the limits on Congress's power under §5 of the Fourteenth Amendment. Second, however dangerous the power to make rules like Miranda might be if they are subject to revision by Congress, the power to make the same rules immune from revision by Congress would be vastly more dangerous.
Amicus parts company with Professor Monaghan when he says constitutional common law could be used "to impose on the states . . . all the best features of the Federal Rules of Criminal Procedure and the Federal Rules of Evidence." Monaghan, supra, 89 Harv. L. Rev., at 43. That would be precisely the supervisory power this Court has always disclaimed. See, e.g., Victor v. Nebraska, 511 U. S. 1, 17 (1994).
The proper line of demarcation is between the scope of the right and the enforcement of the right. This is the same line that marks the limit of Congress's enforcement powers. Congress cannot expand the substantive scope of rights so as to take from states powers left to them by the Constitution. See City of Boerne v. Flores, 521 U. S. 507, 519 (1997). Enforcement may sometimes sweep more broadly than the constitutional provision itself, when necessary to protect a right from being denied in practice through the burden and uncertainty of case-by-case litigation. A ban on all literacy tests for voting, for example, swept out valid tests along with the ones so notoriously used for racial discrimination, yet the ban was valid. See Oregon v. Mitchell, 400 U. S. 112, 283-284 (1970) (Stewart, J., concurring in part and dissenting in part). This Court refers to such legislation as "prophylactic," see Kimel v. Florida Bd. of Regents, 528 U. S. __ (No. 98-791, Jan. 11, 2000) (slip op., at 24), the same word it uses to describe Miranda.
As applied to criminal procedure, enforcement consists of two kinds of rules. Some enforcement rules govern the procedure and evidence by which a court determines whether a violation has occurred, or, in self-incrimination suppression hearings, whether admission of the evidence would be a violation. The other kind of enforcement rule is a rule of remedy. Once a court determines a violation has occurred, what, if anything, does it do about it? Chapman and Mapp v. Ohio, 367 U. S. 643 (1961) are rules of remedy. These two kinds of rules are different from rules defining the scope of the right, such as whether a detention is "reasonable," see, e.g., Illinois v. Wardlow, 528 U. S. __ (No. 98-1036, Jan. 12, 2000), or whether a definition of "reasonable doubt" comports with the Due Process Clause. See, e.g., Victor, supra, 511 U. S., at 15.
The distinction can be illustrated by comparing the law of admissibility of confessions with the law of admissibility of prior convictions. The better practice, it is widely recognized, is to keep the information on prior convictions presented to the jury to the minimum the jury needs to answer the questions before it. See Old Chief v. United States, 519 U. S. 172, 185-186 (1997). This Court can require federal district courts to observe that practice in federal prosecutions, either by supervisory power or by extrapolation from the very general wording of federal rules. See id., at 191-192. However, there is no constitutional prohibition against telling the jury about the priors. See Spencer v. Texas, 385 U. S. 554, 563 (1967). As there is no constitutional violation and no danger of one, this Court has no authority to impose the better practice on the states. See id., at 564. With confessions, the Constitution draws the line at voluntariness. The Constitution itself excludes confessions which are actually over the line, and the federal enforcement power is limited to excluding those which present an unacceptable risk of being over the line. How much risk is acceptable is a policy choice subject to congressional revision.
Miranda thus falls with a narrow definition of the constitutional common law, one that can be adopted without making the federal judiciary a more dangerous branch than it already is. The competing proposition, that this Court should constitutionalize Miranda, involves a far greater danger. Indeed, the very idea that judicial restraint could be proffered as a rationale for this step is remarkable, to say the least. As we discuss in part IV, infra, at 26, Miranda had no basis in precedent or in the text or history of the Constitution. If the Fifth Amendment did not require this formulation when it was adopted, did not require it when it was "incorporated" in the Fourteenth, and did not require it any time prior to 1966, how can it require it today? That would be possible only if this Court had the power to enlarge the scope of constitutional rights based on its own notions of policy, as opposed to construing the Constitution based on its text and history to determine what rights it actually confers. The power to not only promulgate a rule like Miranda but also to place it above congressional revision would be vastly more dangerous than a properly limited common law enforcement power.
A decision upholding 18 U. S. C. §3501 would not be the end of Miranda v. Arizona, 384 U. S. 436 (1966). It would not even be the beginning of the end. It might be the end of the beginning. Cf. W. Churchill, Speech at the Lord Mayor's Day Luncheon (Nov. 10, 1942), quoted in J. Bartlett, Familiar Quotations 746 (15th ed. 1980).
The statute, by its terms, applies only to federal prosecutions. See §3501(a). These are the cases where Miranda was least needed. Federal law enforcement had already adopted a form of warning before Miranda, see 384 U. S., at 483-484, and doubtless will continue giving warnings under the statute. See §3501(b)(3), (4) (warnings are a factor in determining voluntariness); see also Brief for FBI Agents Association as Amicus Curiae. As federal common law, Miranda will continue as controlling precedent in state courts unless and until Congress replaces it with a different rule. Upholding §3501 would make a dramatic change in the legislative landscape, however, by confirming that Congress does have broad power to prescribe the rules of procedure and evidence for the adjudication of self-incrimination claims.
Those who view the American people as knuckle-dragging Neanderthals with contempt for the Bill of Rights recoil in horror at the very suggestion that the people, through the democratic process, might review and revise the inspired wisdom of Miranda. Amicus CJLF has a more optimistic view of our countrymen.
Defendant asserts that Miranda has widespread acceptance beyond the legal culture, Brief for Petitioner 44, i.e., among the public as a whole. To the extent this statement refers to the core Miranda holding--mandatory warnings for suspects actually arrested--we concur. That is why the warnings are in little or no danger from future congressional action. Although Congress could simply make §3501 apply to the states, it is unlikely to do so. The 1968 act was a legislative overreaction to judicial overreaching: Newton's Third Law applied to politics. Now that tempers have cooled for over three decades, a fresh examination by Congress will surely produce a more nuanced approach. The assertions of law enforcement support for giving the warnings, see Brief of Griffin B. Bell et al. as Amici Curiae 8-11, are beside the point.
The government notes, and amicus agrees, that the core Miranda requirement is easily applied. Brief for the United States 34-35. The government further notes, again correctly, that the extensions of Miranda have blurred the bright line that was originally the rule's greatest virtue. See id., at 35-36. Even the original rule loses its bright-line character when applied to persons who claim to have been "in custody," even though not formally arrested, see id., at 36, n. 26, as the present case vividly illustrates. Difficult questions of voluntariness are simply replaced by "murky and difficult questions of when 'custody' begins." Oregon v. Elstad, 470 U. S. 298, 316 (1985).
Surprisingly, the government submits these observations in support of an argument to cast Miranda into constitutional concrete. Amicus submits that these are powerful arguments to return the issue to legislative control. Congress could require the warnings for those formally arrested and then specify in objective terms what other circumstances will trigger the requirement. Congress could reconsider whether an appropriate weighing of the costs and benefits really warrants the rule of Arizona v. Roberson, 486 U. S. 675 (1988), and it could do so free of the baggage of stare decisis.
As another example, the rule of Doyle v. Ohio, 426 U. S. 610 (1976) excludes valid, probative evidence. See Brecht v. Abrahamson, 507 U. S. 619, 628 (1993). It does so even though use of an arrestee's silence does not violate the self-incrimination privilege. Ibid. It does so only because the use contradicts the implied promise of the mandatory warnings. Id., at 628-629. Once the language of the warnings is under legislative control, Congress could deal with this problem by altering the wording of the warnings, thus restoring a valid and valuable source of evidence needlessly excluded by current doctrine.
In Britain, the warnings are established by regulation. See Police and Criminal Evidence Act of 1984, §66, 12 Halsbury's Statutes of England and Wales 873 (4th ed. 1997). The opposite of our Doyle rule is established by statute, see Criminal Justice and Public Order Act of 1994, §34, 17 Halsbury's Statutes, at 278-279, and so the warning was modified to conform. Id., at 279, notes. British arrestees are now warned, "You do not have to say anything. But it may harm your defence if you do not mention when questioned something which you later rely on in court. Anything you do say may be given in evidence." Police and Criminal Evidence Act, Codes of Practice, Code C §10.4 (rev. ed. 1999). This warning, amicus submits, strikes the balance better than Miranda/Doyle. See Caplan, Questioning Miranda, 38 Vand. L. Rev. 1417, 1469 (1985).
A legislatively controlled law of confessions could take better advantage of changing technology. As recording becomes easier and cheaper, a rule to record all station-house interrogations might be a preferable substitute. See Gardner, Section 1983 Actions Under Miranda: A Critical View of the Right to Avoid Interrogation, 30 Am. Crim. L. Rev. 1277, 1287 (1993). To find out whether it is, Congress could authorize a pilot program limited in time and territory, something the judicial branch could never do through legitimate constitutional jurisprudence.
The original Miranda opinion invited legislative participation, see 384 U. S., at 467, but under conditions no responsible legislator could support. The invitation to construct alternatives came with the implied threat to strike them down ex post facto if the Court found they did not provide equally effective protection for the suspect. See ibid.; Gardner, supra, 30 Am. Crim. L. Rev., at 1287. Invalidation of a standard practice is a disaster, requiring an already overburdened system to retry and possibly set free thousands of convicted criminals. Miranda's invitation gave legislatures a choice between a procedure the Court had endorsed and a stroll through a minefield. While a state might add additional procedures on top of Miranda, no responsible legislature or executive could risk experimenting with a substitute under such conditions. Only a recognition of broad congressional authority, without the strict scrutiny implied by Miranda, will invigorate a fresh democratic debate.
The main impediment to legislative action to enforce the Fifth Amendment is not hostility to civil liberties but simply inertia. Criminal law is not high on the legislative priority list, a problem of long standing. See 4 W. Blackstone, Commentaries 4 (1st ed. 1769).(3) In this case, inertia is on Miranda's side. That precedent remains applicable to the states until new congressional action provides a substitute.
It is high time for a fresh examination of the tangled mass of rules and subrules that make up the body of jurisprudence begun by Miranda and extended by cases such as Roberson and Doyle. The legislative branch has the flexibility and adaptability the task requires. It is time to remove the Miranda straitjacket and allow the people's representatives to resolve these issues. Amicus CJLF urges the Court to throw open the doors and let in the fresh air of a robust new round of democratic debate.
|
|
||||
|
|
|
|
||
3. "Were even a committee appointed but once in a hundred years to revise the criminal law, it could not have continued to this hour [1769] a felony without benefit of clergy [i.e., a capital offense], to be seen for one month in the company of persons who call themselves, or are called, Egyptians."